MongoDB Atlas SQL JDBC Driver

The MongoDB Atlas SQL JDBC Driver provides SQL connectivity to MongoDB Atlas for client applications developed in Java.
See the Atlas SQL Documentation for more information.

MongoDB Atlas SQL JDBC is a JDBC Type 4 driver compatible with the JDBC 4.2 specification.

Usage

Prerequisites

MongoDB Atlas SQL JDBC driver requires Java 1.8 or higher.

Download and Install

You can download the precompiled driver (jars) from Maven Central.
Choose jar in the Download dropdown for the lean jar (dependencies not included) or all.jar for the fat jar (dependencies bundled inside the jar).

Connection URL and properties

Connection URL

The connection URL is based on MongoDB connection string and has the jdbc: prefix. The general format for the connection URL is as follows, with items in square brackets ([ ]) being optional:

jdbc:mongodb://[username:password@]host1[:port1][,...hostN[:portN]][/[defaultauthdb][?option1=value1[&option2=value2]...]

For more details :

• URI format: https://www.mongodb.com/docs/manual/reference/connection-string/
• Supported options: https://www.mongodb.com/docs/manual/reference/connection-string/#std-label-connections-connection-options

Notes

• Special characters in the JDBC url have to be URL encoded.
• The driver can not connect to a mongod instance, only to MongoDB Atlas.

Connection Properties

All connection options can also be specified through a Properties object parameter instead of the being directly in the URL. However, if an option is in both the URL and the Properties object, the connection will fail.

In addition to the standard MongoDB connection options there are a number of additional properties specific to the JDBC driver.
These properties can only be specified using an additional Properties object parameter and not in the URL.

Property	Type	Required	Default	Description
database	String	Yes	Null	The name of the database used when querying
loglevel	String	No	OFF	The log level used for logging. Supported levels by increasing verbosity are 'OFF', 'SEVERE', 'WARNING', 'INFO', 'FINE' and 'FINER'
logdir	String	No	Null	The directory to use for log files. If no logging directory is specified, the logs are sent to the console

The following example demonstrates how to open a connection specifying :

• The standard options user and password via a Properties object and ssl and authSource via the URL.
• The JDBC specific options database (mandatory) and loglevel via a Properties object.

         java.util.Properties p = new java.util.Properties();
         p.setProperty("user", "user");
         p.setProperty("password", "foo");
         p.setProperty("database", "test");
         p.setProperty("loglevel", Level.SEVERE.getName());
         System.out.println("Connecting to database test...");
         Connection conn = DriverManager.getConnection("mongodb://mydatalake-xxxx.a.query.mongodb.net/?ssl=true&authSource=admin", p);

Development

Build From Source

To build and test the driver run the following commands from root dir.

To build the lean jar which does not include the dependencies

./gradlew clean build

To run demo locally

./gradlew clean :demo:run

To generate the source Jar

./gradlew clean :sourceJar

You can find the generated jar in build/libs/

To generate the fat Jar which includes all the dependencies

./gradlew clean :shadowJar

You can find the generated jar in build/libs/

To generate the test Jar

./gradlew clean :testJar

You can find the generated jar in build/libs/

To run the unit tests

./gradlew clean test

To fix lint problems

./gradlew spotlessApply

Integration Testing

Integration testing requires a local MongoDB and Atlas Data Lake instance to be running

Environment Variables

ADL_TEST_LOCAL_USER: Local ADL username
ADL_TEST_LOCAL_PWD: Local ADL password
ADL_TEST_LOCAL_AUTH_DB: Local ADL authentication database
HAVE_LOCAL_MONGOHOUSE: "1" if using local mongohouse source
LOCAL_MONGOHOUSE_DIR: Path to local mongohouse source

MDB_TEST_LOCAL_PORT (Optional): Local MongoDB port

To load integration test data

./gradlew runDataLoader

To generate integration test baseline configuration files

./gradlew runTestGenerator

Running Local mongod and Atlas Data Lake

run_adl.sh is a helper script that will start a local mongod and Atlas Data Lake instance, used for integration testing.

Start

./resources/run_adl.sh start

Stop

./resources/run_adl.sh stop

Skip

Use the SKIP_RUN_ADL option to skip the start and stop operations for those managing their own local instances.

export SKIP_RUN_ADL=1